Low Computational Complexity, High Coding Efficiency Intra Prediction for HEVC

碩士 === 國立中央大學 === 通訊工程學系 === 105 === High efficiency video coding (HEVC) is the latest standard for video compression, which can achieve significant improvements on coding efficiency. HEVC adopted quad-tree based coding unit (CU) which provides four kinds of block size for characteristics of video....

Full description

Bibliographic Details
Main Authors: Han-Yuan Hsu, 許瀚元
Other Authors: Yin-Yi Lin
Format: Others
Language:zh-TW
Published: 2016
Online Access:http://ndltd.ncl.edu.tw/handle/bbse78
Description
Summary:碩士 === 國立中央大學 === 通訊工程學系 === 105 === High efficiency video coding (HEVC) is the latest standard for video compression, which can achieve significant improvements on coding efficiency. HEVC adopted quad-tree based coding unit (CU) which provides four kinds of block size for characteristics of video. CU contains Prediction Unit (PU) and Transform Unit (TU). To improve intra coding efficiency, using 35 prediction modes in PU. However, HEVC encoder complexity is tremendously increased. Hence, in this paper, we proposed a SVM based fast intra CU depth decision algorithm and utilize the minimum Rate-distortion Cost (RD Cost) to develop the criterion of CU early termination under SVM based algorithm. HEVC adopt RMD algorithm which select 8, 8, 3, 3, 3 modes for candidates list in different PU size. In our intra mode decision, the modes with larger SATD Cost are removed from candidate list of RDO process. The threshold is adaptive by block size and QP. The experimental results show that our proposed algorithm achieves 22% time saving on average, with 0.09% BD-rate increase.